Load pre-processed single reference datasets. More...

Classes
struct	LoadLabelsOptions
	Options for loading labels. More...

struct	LoadMarkersOptions
	Options for loading markers. More...

struct	LoadRankingsOptions
	Options for loading rankings. More...

Typedefs
template<typename Value_ = double, typename Index_ = int, typename Rank_ = int>
using	RankMatrix = tatami::DenseMatrix< Value_, Index_, std::vector< Rank_ > >

Functions
template<typename Label_ = singlepp::DefaultLabel>
std::vector< Label_ >	load_labels_from_text_file (const char *path, const LoadLabelsOptions &options)

template<typename Label_ = singlepp::DefaultLabel>
std::vector< Label_ >	load_labels_from_gzip_file (const char *path, const LoadLabelsOptions &options)

template<typename Label_ = singlepp::DefaultLabel>
std::vector< Label_ >	load_labels_from_zlib_buffer (const unsigned char *buffer, size_t len, const LoadLabelsOptions &options)

template<typename Index_ = singlepp::DefaultIndex>
singlepp::Markers< Index_ >	load_markers_from_text_file (const char *path, const LoadMarkersOptions &options)

template<typename Index_ = singlepp::DefaultIndex>
singlepp::Markers< Index_ >	load_markers_from_gzip_file (const char *path, const LoadMarkersOptions &options)

template<typename Index_ = singlepp::DefaultIndex>
singlepp::Markers< Index_ >	load_markers_from_zlib_buffer (const unsigned char *buffer, size_t len, const LoadMarkersOptions &options)

template<typename Value_ = singlepp::DefaultValue, typename Index_ = singlepp::DefaultIndex>
RankMatrix< Value_, Index_ >	load_rankings_from_text_file (const char *path, const LoadRankingsOptions &options)

template<typename Value_ = singlepp::DefaultValue, typename Index_ = singlepp::DefaultIndex>
RankMatrix< Value_, Index_ >	load_rankings_from_gzip_file (const char *path, const LoadRankingsOptions &options)

template<typename Value_ = singlepp::DefaultValue, typename Index_ = singlepp::DefaultIndex>
RankMatrix< Value_, Index_ >	load_rankings_from_zlib_buffer (const unsigned char *buffer, size_t len, const LoadRankingsOptions &options)

template<typename Value_ , typename Index_ , typename Rank_ , typename Label_ >
void	verify (const RankMatrix< Value_, Index_, Rank_ > &rankings, const std::vector< Label_ > &labels, const singlepp::Markers< Index_ > &markers)

Detailed Description

Load pre-processed single reference datasets.

Typedef Documentation

◆ RankMatrix

template<typename Value_ = double, typename Index_ = int, typename Rank_ = int>

using singlepp_loaders::RankMatrix = typedef tatami::DenseMatrix<Value_, Index_, std::vector<Rank_> >

Matrix of ranks as a dense column-major matrix. Each column corresponds to a sample while each row corresponds to a feature. Each column contains the ranked expression values for all features.

Template Parameters

Value_	Numeric type for data in the matrix interface.
Index_	Integer type for indices in the matrix interface.
Rank_	Integer type for the ranks being stored.

Function Documentation

◆ load_labels_from_gzip_file()

template<typename Label_ = singlepp::DefaultLabel>

std::vector< Label_ > singlepp_loaders::load_labels_from_gzip_file	(	const char *	path,
		const LoadLabelsOptions &	options
	)

Template Parameters

Label_ Integer type for the label identity.

Parameters

path	Path to a Gzip-compressed file containing the labels.
options	Options for reading the labels.

Returns: Vector containing the label index for each reference profile.

See load_labels_from_text_file() for details about the format.

◆ load_labels_from_text_file()

template<typename Label_ = singlepp::DefaultLabel>

std::vector< Label_ > singlepp_loaders::load_labels_from_text_file	(	const char *	path,
		const LoadLabelsOptions &	options
	)

Template Parameters

Label_ Integer type for the label identity.

Parameters

path	Path to a text file containing the labels.
options	Options for reading the labels.

Returns: Vector containing the label index for each reference profile.

The file should contain one line per profile, containing an integer label index for that profile. Label indices refer to another array containing the actual names of the labels (see load_label_names_from_text_file()). The total number of lines should be equal to the number of profiles in the dataset. The file should not contain any header.

◆ load_labels_from_zlib_buffer()

template<typename Label_ = singlepp::DefaultLabel>

std::vector< Label_ > singlepp_loaders::load_labels_from_zlib_buffer	(	const unsigned char *	buffer,
		size_t	len,
		const LoadLabelsOptions &	options
	)

Template Parameters

Label_ Integer type for the label identity.

Parameters

[in]	buffer	Pointer to an array containing a Zlib/Gzip-compressed string of labels.
	len	Length of the array for `buffer`.
	options	Options for reading the labels.

Returns: Vector containing the label index for each reference profile.

See load_labels_from_text_file() for details about the format.

◆ load_markers_from_gzip_file()

template<typename Index_ = singlepp::DefaultIndex>

singlepp::Markers< Index_ > singlepp_loaders::load_markers_from_gzip_file	(	const char *	path,
		const LoadMarkersOptions &	options
	)

Template Parameters

Index_ Integer type for the marker indices.

Parameters

path	Path to a Gzip-compressed file containing the marker lists.
options	Further options for reading.

Returns: A Markers object containing the markers from each pairwise comparison between labels.

See load_markers_from_text_file() for details about the format.

◆ load_markers_from_text_file()

template<typename Index_ = singlepp::DefaultIndex>

singlepp::Markers< Index_ > singlepp_loaders::load_markers_from_text_file	(	const char *	path,
		const LoadMarkersOptions &	options
	)

Template Parameters

Index_ Integer type for the marker indices.

Parameters

path	Path to a text file containing the marker lists.
options	Further options for reading.

Returns: A Markers object containing the markers from each pairwise comparison between labels.

The file should contain one line per pairwise comparison between labels. Each line should at least 3 tab-delimited fields - the index of the first label, the index of the second label, and then the indices of the features selected as marker genes for the first label relative to the second. Any (non-zero) number of marker indices may be reported provided they are ordered by marker strength. The total number of lines in this file should be equal to the total number of pairwise comparisons between different labels, including permutations.

◆ load_markers_from_zlib_buffer()

template<typename Index_ = singlepp::DefaultIndex>

singlepp::Markers< Index_ > singlepp_loaders::load_markers_from_zlib_buffer	(	const unsigned char *	buffer,
		size_t	len,
		const LoadMarkersOptions &	options
	)

Template Parameters

Index_ Integer type for the marker indices.

Parameters

[in]	buffer	Pointer to an array containing a Zlib/Gzip-compressed string containing the marker lists.
	len	Length of the array for `buffer`.
	options	Further options for reading.

Returns: A Markers object containing the markers from each pairwise comparison between labels.

See load_markers_from_text_file() for details about the format.

◆ load_rankings_from_gzip_file()

template<typename Value_ = singlepp::DefaultValue, typename Index_ = singlepp::DefaultIndex>

RankMatrix< Value_, Index_ > singlepp_loaders::load_rankings_from_gzip_file	(	const char *	path,
		const LoadRankingsOptions &	options
	)

Template Parameters

Value_	Numeric type for data in the matrix interface.
Index_	Integer type for indices in the matrix interface.

Parameters

path	Path to a Gzip-compressed file containing the ranking matrix.
options	Options for reading the rankings.

Returns: A RankMatrix containing the feature rankings for each reference profile. Each column corresponds to a reference profile while each row corresponds to a feature.

See load_rankings_from_text_file() for details about the format.

◆ load_rankings_from_text_file()

template<typename Value_ = singlepp::DefaultValue, typename Index_ = singlepp::DefaultIndex>

RankMatrix< Value_, Index_ > singlepp_loaders::load_rankings_from_text_file	(	const char *	path,
		const LoadRankingsOptions &	options
	)

Template Parameters

Value_	Numeric type for data in the matrix interface.
Index_	Integer type for indices in the matrix interface.

Parameters

path	Path to a text file containing the ranking matrix.
options	Options for reading the rankings.

Returns: A RankMatrix containing the feature rankings for each reference profile. Each column corresponds to a reference profile while each row corresponds to a feature.

The file should contain one line per reference profile, with the total number of lines equal to the number of profiles in the dataset. Each line should contain the rank of each feature's expression within that profile, separated by commas. The number of comma-separated fields on each line should be equal to the number of features. Ranks should be strictly integer - tied ranks should default to the minimum rank among the index set of ties.

◆ load_rankings_from_zlib_buffer()

template<typename Value_ = singlepp::DefaultValue, typename Index_ = singlepp::DefaultIndex>

RankMatrix< Value_, Index_ > singlepp_loaders::load_rankings_from_zlib_buffer	(	const unsigned char *	buffer,
		size_t	len,
		const LoadRankingsOptions &	options
	)

Template Parameters

Value_	Numeric type for data in the matrix interface.
Index_	Integer type for indices in the matrix interface.

Parameters

[in]	buffer	Pointer to an array containing a Zlib/Gzip-compressed string containing the ranking matrix.
	len	Length of the array for `buffer`.
	options	Options for reading the rankings.

Returns: A RankMatrix containing the feature rankings for each reference profile. Each column corresponds to a reference profile while each row corresponds to a feature.

See load_rankings_from_text_file() for details about the format.

◆ verify()

template<typename Value_ , typename Index_ , typename Rank_ , typename Label_ >

void singlepp_loaders::verify	(	const RankMatrix< Value_, Index_, Rank_ > &	rankings,
		const std::vector< Label_ > &	labels,
		const singlepp::Markers< Index_ > &	markers
	)

This checks the consistency of all components of the reference dataset after they have been loaded:

The number of columns of rankings should be equal to the length of labels.
The number of labels is defined as the largest value in labels plus 1.
The length of markers and the number of inner vectors in each element of markers should be equal to the number of labels.
Each gene index in markers should be non-negative and less than the number of rows in rankings.

Template Parameters

Value_	Numeric type for data in the matrix interface.
Index_	Integer type for indices in the matrix interface.
Rank_	Integer type for the ranks being stored.
Label_	Integer type for the labels.

Parameters

rankings	The matrix produced by `load_rankings_from_text_file()` or related functions.
labels	Vector of labels produced by `load_labels_from_text_file()` or related functions.
markers	List of markers produced by `load_markers_from_text_file()` or related functions.

Classes

Typedefs

Functions

Detailed Description

Typedef Documentation

◆ RankMatrix

Function Documentation

◆ load_labels_from_gzip_file()

◆ load_labels_from_text_file()

◆ load_labels_from_zlib_buffer()

◆ load_markers_from_gzip_file()

◆ load_markers_from_text_file()

◆ load_markers_from_zlib_buffer()

◆ load_rankings_from_gzip_file()

◆ load_rankings_from_text_file()

◆ load_rankings_from_zlib_buffer()

◆ verify()